Wednesday, January 30, 2013

Substance versus Procedure

Posted by Sherry F. Colb

In my Verdict column for this week, I discuss Evans v. Michigan, a Double Jeopardy case pending in the U.S. Supreme Court.  The issue in Evans is whether the Fifth Amendment right against being tried more than once for the same crime precludes retrial when a judge, after the close of the State's evidence, grants the defendant's motion for a directed verdict on the basis of the judge's having mistakenly added an additional, non-existent element to the crime to be proved.  In my column, I discuss the issue in Evans in the light of the Court's decision in Blueford v. Arkansas last term.

In this post, I want to consider the larger question of whether substance is more or less important to a just system than is process.  The Double Jeopardy Clause provides a primarily process-oriented right to criminal defendants, although it may serve substantive values, as I suggest in my column, by preventing an oppressive government agent from repeatedly ignoring proper acquittals until a convicting jury provides the "desired" guilty verdict.  The trial itself is a procedural mechanism, but the goal is much more straightforwardly substantive -- to find out the truth of what occurred so that the guilty may be punished, while the innocent are spared.

The criminal trial is not, however, exclusively concerned about the truth.  For one thing, the heavy burden that is constitutionally placed on the government's shoulders reflects a willingness to tolerate inaccuracy in the form of acquittal of many guilty people in the service of minimizing the risk of convicting one innocent person.  More directly threatening to the truth are evidentiary privileges that keep reliable, relevant evidence from the fact-finder, to serve some extrinsic interest (such as protecting the sanctity and security of private marital communications) and exclusionary criminal procedure rules (such as the Fourth Amendment exclusionary rule).  Despite these compromises, however, we would not necessarily say that the trial itself de-prioritizes truth, but only that other values make their way into a process that is itself primarily concerned with uncovering the truth about whether the defendant did or did not commit the crime charged.

I sometimes wonder, though, whether even that much is accurate.  The trial is, as a fact-finding procedure, surely superior to old practices such as trial by ordeal.  But is it all about learning the truth?  Years ago, Professor Dorf asked his constitutional law students the following question:  if you could replace trial by jury with a machine that could tell us, with 99.999% certainty, whether or not the defendant is guilty, would you do it?  My recollection (admittedly hearsay, another problem in a trial setting) is that many, if not most, students rejected the machine in favor of the jury, notwithstanding the acknowledged fact that juries are not nearly as accurate as the hypothetical machine.  For those students (and perhaps for some readers), the trial is preferable to even a more-accurate device for finding out the truth.  But how could that be?

One possibility is that students were "fighting the hypo" and imagining that the device wasn't really as accurate as stipulated, and that if this were so, it would be nearly impossible to know about the flaw because the device is, by definition, opaque.  If students had truly believed in the accuracy of the device, perhaps they would have come out the other way.

On the other hand, because the jury deliberates in private, we don't really know in any case (or, for that matter, in the run of cases), whether juries are getting things right.  We have some sobering evidence suggesting they are not, in the form of DNA exonerations, sometimes many years after eye-witness identifications led to convictions.  The absolute preference for juries therefore does seem to prioritize something over the accuracy of outcomes, but what?

I would tentatively identify the "something" as the individual's access to a process in which a group of presumptively uncorrupted individuals who take their job seriously and who themselves could some day be subject to the same process listen to everyone involved (including the defendant, if he or she chooses to testify) and comes to an unbiased conclusion after deliberation.  There is something familiar and comforting about the jury process, particularly when compared to a foreign-looking device.  For society at large, it "feels better" to have juries deciding guilt and innocence, much in the way that it might feel better to talk to an actual person, rather than a computer robot, when calling a business on the telephone, quite apart from relative efficiency and efficacy.

If I am right about this assessment, however, then there is at least one problem.  The innocent person on trial for a crime would probably choose the most accurate fact-finding device rather than risk having a human jury make a mistake.  To the extent that many of us prefer a jury, then, even with the accuracy stipulation, I worry that it is because the jury makes us feel good as spectators to criminal justice, regardless of its potential inferiority as a truth-seeking device for innocent defendants.  In other words, I wonder whether a preference for process over substance in this context (and maybe in others) does more for the perceived legitimacy of the system (and the comfort with which those who confidently predict they will never fall prey to it) than it does for its efficacy in protecting the innocent from falsely being convicted.  In that sense, perhaps, the trial of today does bear some resemblance to trials by ordeal of yesteryear.


14 comments:

Michael C. Dorf said...

Your recollection is more or less accurate with a couple of small caveats: 1) I used to ask the question of my civil procedure students, not my con law students, as an introduction to a unit on procedural fairness; and 2) I would usually start with some social science statistics about the accuracy of juries and then progressively make the hypothetical machine more and more accurate. Nobody would pick the machine when it was about the same (at truth detection) as the jury, and I'd need to get to something very close to 100% before even getting a few students to say they would take the machine.

When I swapped federal courts for civil procedure six years ago, I stopped asking the question so I don't have new data, but I doubt much has changed.

Paul Scott said...

Is it possible that people prefer the ability of the jury to make value judgments about the law?

My issue with the machine is not its accuracy, but its inability to ignore the instructions of the judge and do what the collection of humans deems right.

This may be, in fact likely is, a reflect of my own potentially extreme view, but I, as a juror, would never listen to instructions like "you are only a finder of fact and must not take into account considerations of punishment when reaching your decision." And, yes, that is because I feel, justified or not, that I am better able to decide what is right and wrong than is our political process that generated the criminal laws.

I think the issue is not simply a matter of comfort with the familiar, but is a matter of comfort with justice. Factually, on aggregate, the machine you discuss with the laws we have now, would likely still produce greater overall justice (in that whatever failings it has from being incapable of not following instructions will be overwhelmingly made up by replacing factual inaccuracies of the jury system), so I would take the machine. In truth, I would probably be willing to replace a jury with lie detectors if such a thing were economically possible (it would be in some cases, but not in most).

But my objection or discomfort with the machine is unrelated to theatrics.

Mark Regan said...

The other thing is that in a criminal trial accuracy sometimes takes a back seat to the community's sense of what ought to be done. Many years ago, in a rural Alaska community, there was an intra-family shooting where the 99.9% accurate machine would have come back with second-degree murder. The jury thought about the impact on the family of semi-permanently incarcerating the defendant and came back with manslaughter. Justice was done.

Sam Rickless said...

Mike -- The experiment with your students, though admittedly not designed to be *scientific*, is potentially flawed as a device for registering what knowledgeable people on reflection would prefer as a system of guilt detection. The problem has to do with the context in which the relevant questions are asked. If you start with a machine that is as accurate as a jury and then pose successive hypos in which the machine is progressively more accurate, respondents get caught in a kind of Sorites bind. Aware that they prefer the jury over the machine in Hypo #1, when you make the machine slightly more accurate in Hypo #2, respondents don't see the difference in accuracy between the Hypo #1 machine and the Hypo #2 machine as relevant. Understandably, then, they are likely to prefer jury over machine in Hypo #2. As you move through the Hypos, respondents keep saying the same thing (because Hypo #n-1 is relevantly indistinguishable from Hypo #n), until they reach 99%, which is when the enormity of what they have been brought to accept jogs (some of) them out of their Sorites-induced stupor. It would be interesting to see what sort of reaction you would get if you took two groups of respondents: to one group, you would ask the question whether you prefer a machine of 75% accuracy to a jury; to the other, you would ask the question whether you prefer a machine of 99.99% accuracy to a jury. My prediction is that the results will be different if the Sorites effect is avoided. Another interesting experiment would be to *start* with a machine that is 100% accurate, and then move the accuracy percentage progressively lower, all the while asking respondents whether they would prefer machine to jury. Here I predict that many more respondents would prefer a 99% accurate machine after being primed with an example containing a 100% accurate machine than prefer a 99% accurate machine after having been primed with examples containing machines with much lower levels of accuracy.

Michael C. Dorf said...

Sam: I'm sufficiently disorganized in my teaching that I have posed the question different ways. My recollection is that even when I started with the machine being much more accurate, very few students preferred it.

t jones said...

Or, law students prefer jury trials because their future selves' jobs would be threatened by a machine which could reach the right result, which would almost assuredly require less input from attorneys.

Sam Rickless said...

The idea that you would be "sufficiently disorganized in your teaching", Mike, strikes me as too absurd to require comment....

But, since you insist...;-)...I will ask my 324 "ethics and society" students the very same question from the other direction on Wednesday when I see them next. (On Monday, there will be a guest lecture.) That will give us one more data point. Of course, *I* am sufficiently disorganized in *my* teaching that I will surely forget to ask my students what they think. So don't hold me to it!

Michael C. Dorf said...

Sam: That's great. I look forward to the results of your survey!

Andrew Pollis said...

I think the problem with the machine hypo is that it works only if the defendant's guilt is objectively true or false (e.g., did Joe shoot the gun that killed Sam?). It doesn't work, as some of the prior comments have suggested, when guilt hinges on a subjective value judgment of some kind or when state of mind is an issue (e.g., did Joe believe that Sam was about to shoot HIM, such that he was acting in self-defense?). It gets even harder to dispense with the human factor if we insert an objective element on top of the subjective (e.g., did Joe REASONABLY believe Sam was about to shoot him?). I don't think, where these kinds of questions are involved, that process can be meaningfully distinguished from substance.

Andrew Pollis said...

I was just coincidentally reading a 1979 article by Charles Nesson on the dangers of permissive inferences in criminal cases. He uses an interesting hypothetical that demonstrates the law's preference for jury verdicts based on subjective evaluations of credibility, even when less mathematically likely to be accurate. See Charles R. Nesson, Reasonable Doubt and Permissive Inferences: The Value of Complexity, 92 HARV. L. REV. 1187, 1192-94 (1979).

Cicy said...

to the other, you would ask the question whether you prefer a machine of 99.99% accuracy to a jury. My prediction is that the results will be different if the Sorites effect is avoided. Another interesting experiment would be to *start* with a machine that is 100% accurate, and then move the accuracy percentage progressively lower, all the while asking respondents whether they would prefer machine to jury. Here I predict that many more respondents would prefer a 99% accurate machine after being primed with an example containing a 100% accurate machine than prefer a 99% accurate machine after having been primed with examples containing machines with much lower levels of accuracy.

Fifa 14 Ultimate Team Coins -FIFA 2014 World Cup Brazil Coins PS3
LOL elo boost

marly aaran said...

Great tips and very easy to understand. This will definitely be very useful for me when I get a chance to start my blog. Thanks for sharing such a great content with us.
devthroughsport.org |

amine lahragui said...


thanks so much for that great blog and thanks also for accepting my links thanks
طريقة عمل الدونات طريقة عمل البان كيك طريقة عمل الكنافة طريقة عمل البسبوسة طريقة عمل الكيك طريقة عمل عجينة البيتزا فوائد القرفه

Jessica Colin said...

Today I said goodbye to my first Hospice patient. She would have been 100 in December. They work hard to match up volunteers with patients, and they did a fantastic job with us. We bonded quickly and talked about everything, from how short Tom Cruise is to how worried her five-year-old self was that her
http://www.girlsky.net |